Re: ECI 2002V or 2002R?
Re: ECI 2002V or 2002R?
- Subject: Re: ECI 2002V or 2002R?
- From: Graeme Gill <email@hidden>
- Date: Mon, 30 Sep 2002 12:08:22 +1000
Henrik Holmegaard wrote:
>
IMO a standards-based test chart has advantages over non-standard test
>
charts, and the ECI2002 test chart which is an IT8.7-3 super-set should
>
where possible be used in place of the IT8.7-3 which has always
>
attracted criticism for the way it samples a print color space.
Although the ECI2002 is definitely an improvement over the IT8.7-3,
it still has the problem that all of device values line up
(ie. most sampling values of CMYK are all taken from the same
small number of values, 0, 10, 20, 30, 40 55, 70, 85, 100%,
with a few patches using an extra 11 values.)
This causes a problem in creating high quality profiles,
because most of the possible device channel values aren't being
sampled at all, and those that are, are being sampled
redundantly. This makes the creation of detailed device
linearisation curves quite difficult or impossible.
Using a rectangular grid (as the ECI and IT8 charts do) is an
inefficient way to sample a 3D space. Hexagonal sampling would
be nearly 50% more efficient for instance. The ECI chart is good
in varying its sampling density through the color space, to make
the patches more evenly spread perceptually, but a more stochastic
sampling pattern would avoid mathematical "aliasing" problems
in constructing the profile interpolation structures.
So, I think that the ECI2002 is the best standard chart so far,
but isn't as good as a well made (ie. state of the art) custom test
chart.
>
The iCColor x-y chart reader supports a scrambled ECI2002 layout. When
>
the slowest, two part version is measured on OS 10.2 it takes 16
>
minutes for 1500 patches. The one part version is a couple of minutes
>
faster. The SpectroScan x-y chart reader supports a systematic and a
>
scrambled layout.
It's interesting that in reading 1500 patches, the Gretag Spectroscan is
more convenient (set it going and walk away), but the Xrite DTP41 is
faster.
Graeme Gill.
_______________________________________________
colorsync-users mailing list | email@hidden
Help/Unsubscribe/Archives:
http://www.lists.apple.com/mailman/listinfo/colorsync-users
Do not post admin requests to the list. They will be ignored.